摘要
讨论了一种直接甲醇燃料电池基于机理模型和神经网络模型的混合模型构建方法,利用人工神经网络的非线性逼近能力,对机理模型的不精确性进行有效补偿。混合模型中的神经网络模型根据输入变量和补偿量训练后与机理模型相结合,对电池电压提供了良好的近似预测。
This article discusses a method of direct methanol fuel cell hybrid modeling based mechanism model and artificial neural network model. The artificial neural network of the hybrid model has the capacity of non-linear approximation, and it can give an effective compensation for the inaccurate of mechanism model. The neural network model can give a good prediction of the output variables after training with input variables and compensation combined with mechanism model.
出处
《电子技术应用》
北大核心
2010年第3期67-70,74,共5页
Application of Electronic Technique
基金
河南省科技创新人才计划(项目编号:84200510009)